1. Field of the Invention
The present invention relates to a stereophonic signal processing apparatus, and more particularly, to a stereophonic signal processing apparatus having a vocal cancel function to eliminate vocal signals included in left and right input signals.
2. Description of the Related Art
Conventionally, a circuit as shown in FIG. 6 is used to cancel vocal components (vocal signals) from left and right input stereophonic signals mixedly including vocal and instrumental plays and to obtain a instrumental play signal.
This circuit has a left low-pass filter 11 that extracts a left low-frequency signal and a left high-pass filter 12 that extracts a left high-frequency signal, from a left input signal (Lin), a right low-pass filter 21 that extracts a right low-frequency signal and a right high-pass filter 22 that extracts a right high-frequency signal, from a right input signal (Rin), subtraction means 31 that performs subtraction between the right input signal (Rin) and the left input signal (Lin) to generate a vocal elimination signal, left addition means 13 that adds the left low-frequency signal and the left high-frequency signal to the vocal elimination signal to generate a left output signal (Lout), and right addition means 23 that adds the right low-frequency signal and the right high-frequency signal to the vocal elimination signal to generate a right output signal (Rout).
First, the vocal elimination signal where the vocal signal precisely located around the center of the stereophonic signal is canceled, is generated by subtraction between the left input signal and the right input signal at equal ratio, by the subtraction means 31. In the vocal elimination signal, though the vocal signals are canceled, the signal components with different phases (play signals) from the left and right signals remain. Note that the vocal elimination signal is a monophonic signal.
Then, the left and right filters, having characteristics as shown in FIG. 7(a), respectively extract a low-frequency signal and a high-frequency signal except an intermediate frequency including a vocal signal in advance. The low-frequency signals and the high-frequency signals are stereophonic signals.
Then, from the vocal elimination signal as a monophonic signal, to obtain the stereo effect and to enhance the low-frequency and high-frequency regions, stereophonic signals, where all the frequencies are included and the vocal signals are eliminated, are generated by adding the left and right respective low-frequency signals and high-frequency signals.
In the above circuit, signals of musical instruments precisely located around the center of the stereophonic signals are deleted with the vocal signals by the subtraction, however, a sufficient level output can be obtained by passing the signals through the low-pass filter or high-pass filter (see FIG. 7A).
However, in a signal of an instrument having frequency components as shown in FIG. 7B, although the signal has frequency components from a fundamental f0 to f2, frequency components around the frequency component f1 are cut by the low-pass filter. That is, frequency components in the region A01 remain but frequency components in the region A12 are eliminated.
As these frequency components are eliminated, for example, an attack sound of an instrument, such as, a bass drum is changed to a base-like sound. That is, the sound of an original instrument changes as if it is a sound of another instrument.
Further, a similar phenomenon occurs around a cutoff frequency of the high-pass filter in the sound of an instrument having a peak around the frequency.
Then, to avoid the above problem, the characteristics of the low-pass filter or the high-pass filter may be milder. However, such arrangement has a new problem in that the vocal signal to be eliminated passes through the low-pass filter or the high-pass filter.
As a result, since it is necessary to place importance on the performance to eliminate the vocal signals, the change of instrumental sound is not especially considered.
Accordingly, the present invention has an object to realize a stereophonic signal processing apparatus capable of reproducing sounds of musical instruments while effectively eliminating vocal signals.
The invention is embodied in a stereophonic signal processing apparatus for eliminating vocal signals contained in right and left input signals, said apparatus comprising; right filter means for extracting a right low-frequency signal and a right high-frequency signal from a right input signal, left filter means for extracting a left low-frequency signal and a left high-frequency signal from a left input signal, a subtraction means for subtracting a right input signal and a left input signal and thereby generating a vocal elimination signal, a right addition means for adding a right low-frequency signal and a right high-frequency signal from said right filter means to a vocal elimination signal from said subtraction means and thereby generating a right output signal, a left addition means for adding a left low-frequency signal and a left high-frequency signal from said left filter means to a vocal elimination signal from said subtraction means and thereby generating a left output signal, and control means for changing characteristics of both of said right filter means and said left filter means according to at least one of a right input signal and a left input signal.
Note that it is desirable that the control means performs a control to reduce attenuation of signals other than the vocal signal contained in at least one of the right input signal and the left input signal by detecting a rising of a waveform of at least one of the right input signal and the left input signal or detecting a frequency component included in at least. one of the right input signal and the left input signal.
Thus, by changing both characteristics of the right filter means and the left filter means, the attenuation of instrumental sound by the filter means to enhance low-frequency and high-frequency regions can be reduced while the vocal signals are effectively eliminated, and reproduction in high fidelity can be made.
The invention is further embodied in a stereophonic signal processing apparatus for eliminating vocal signals contained in right and left input signals, said apparatus comprising; right filter means for extracting a right low-frequency signal and a right high-frequency signal from a right input signal, left filter means for extracting a left low-frequency signal and a left high-frequency signal from a right input signal, a subtraction means for subtracting a right input signal and a left input signal and thereby generating a vocal elimination signal, a right addition means for adding a right low-frequency signal and a right high-frequency signal from said right filter means to a vocal elimination signal from said subtraction means and thereby generating a right output signal, a left addition means for adding a left low-frequency signal and a left high-frequency signal from said left filter means to a vocal elimination signal from said subtraction means and thereby generating a left output signal, and control means for adjusting, in signal level, a right input signal and a left input signal to be subtracted by said subtraction means according to at least one of a right input signal and a left input signal.
Note that it is desirable that the control means performs a control to reduce attenuation of signals other than the vocal signal contained at least one of the right input signal and the left input signal by detecting a rising of a waveform of at least one of the right input signal and the left input signal or detecting a frequency component contained in at least one of the right input signal and the left input signal.
By adjusting signal levels before subtraction by the subtraction means, the attenuation of instrumental sound by the subtraction means can be reduced, and reproduction in high fidelity can be made.